Licarin B from Myristica fragrans improves insulin sensitivity via PPARγ and activation of GLUT4 in the IRS-1/PI3K/AKT pathway in 3T3-L1 adipocytes

Abstract

Peroxisome proliferator-activated receptors (PPARs) are ligand-activated transcription factors regulating lipid and glucose metabolism. The objective of the present study is to characterize the cellular effect of Licarin B (LB), a neolignan isolated from Myristica fragrans on the PPARγ and insulin signaling pathways in 3T3-L1 preadipocytes. The molecular mechanism of action of LB on PPARγ and insulin signaling pathways were studied using in vitro and in silico methods. Functional activation of PPARγ in vitro was confirmed by 3T3-L1 preadipocyte differentiation, regulation of target genes and protein expression. LB caused triglyceride accumulation during adipogenesis but significantly less compared to rosiglitazone (RG), a PPARγ full agonist. In in vitro time-resolved fluorescence resonance energy transfer-based competitive binding assay, LB showed an IC₅₀ value of 2.4 μM whereas for RG and GW9662 it was 57.96 nM and 18.68 nM respectively. Virtual screening of LB with PPARγ showed hydrophobic interactions with a binding energy of −9.36 kcal mol⁻¹. Interestingly enough LB improved insulin sensitivity by up regulating the GLUT4 expression and translocation via IRS-1/PI3K/AKT pathway, enhanced adiponectin secretion and modulated mRNA expression profile of PPARγ target genes C/EBPα, IRS-2, and LPL significantly compared to RG. This scientifically validate LB as a promising bioactive for insulin resistance and associated complications through its partial PPARγ activity.

---

1. Introduction

Type 2 diabetes mellitus (T2DM) is a worldwide threat that has been labeled as a great challenge to human health in the 21^st century. The total number of people with diabetes is estimated to rise from 382 million to 582 million by 2035.^1,2 Based on the current knowledge of the pathophysiology of insulin resistance and T2DM, various pharmacological and non-pharmacological interventions have been developed to improve glycemic control and prevent diabetes complications. The rising trend in diabetes-related complications suggests that the current prognosis for diabetes is not sufficient and use of supplementary treatment, including functional foods and nutraceuticals, could increase the effectiveness of diabetes management. Natural products from plants have been used in traditional medicine to treat diabetes for thousands of years.^3,4 Although plant extracts have been found to be effective therapeutics, more focus on research is needed for the identification of active molecules and their mode of action. Some of the compounds with antidiabetic activity are alkaloids, saponins, xanthones, flavonoids and nonstarch polysaccharides.⁵ Despite the long list of active principles with experimentally proven antidiabetic activity, to date, metformin is the only drug approved for the treatment of T2DM derived from a plant. Therefore, the identification of active compounds from plants and to delineate their modes of action is an important issue for the discovery of new antidiabetic drugs.

Several mechanisms of antidiabetic action have been proposed for bioactive compounds from traditionally used medicinal plants^6,7 and some hypotheses relate their effects to the increase of the insulin-sensitivity mediated glucose uptake. One target of interest for antidiabetic drugs is peroxisome proliferator-activated receptor gamma (PPARγ). PPARγ is a member of the nuclear receptor superfamily that regulates the gene expression of proteins involved in the control of glucose, lipid and protein metabolism.⁸ Indeed, the importance of PPARγ in regulating insulin sensitivity has inspired research groups to devote many efforts toward developing PPARγ agonists, which could be of therapeutic use for diabetic patients.⁹ Thiazolidinediones (TZDs), an important class of synthetic agonists of PPARγ are antidiabetic agents, currently employed for the treatment of T2DM that target to improve insulin sensitivity. In spite of the clinical benefit of these drugs, the use of TZDs has been associated with adverse effects like weight gain, renal fluid retention and cardiotoxicity.¹⁰ Therefore, new PPARγ ligands with enhanced therapeutic efficacy and reduced adverse effects are required. A promising new class of such ligands is selective PPARγ modulators (i.e., SPPARγMs) or partial PPARγ agonists. These compounds behave as partial agonists of PPARγ and exhibit different binding properties in comparison to full agonists.¹¹ Various natural products and plant extracts have been found to increase insulin-stimulated glucose uptake by modulating the action of PPARγ with little or no effect on adipocyte differentiation.⁷ Thus, PPARγ partial agonists from natural extracts are promising candidates for the treatment of T2DM.

Myristica fragrans Houtt (nutmeg) is an aromatic evergreen tree of the family Myristicaceae.¹² Nutmeg, the actual seed of the tree, is an important nutraceutical, used to treat colds, fever, catarrh, general respiratory ailments, digestion and skin diseases like scabies.¹³ It is also used in pickles and as spices for better digestibility of food in India. In controlled laboratory studies, M. fragrans has been shown to possess insulin-like, antibacterial, antioxidant and also insecticidal properties.^14–16 Since M. fragrans is reported to have antidiabetic effect,¹⁷ it was our interest to check whether Licarin B (LB) isolated from M. fragrans, exerts an effect on any of the molecular pathways relevant to the pathophysiology of diabetes. Keeping this in mind the aim of our present work is to elucidate the mechanism of action of LB on PPAR γ and insulin signaling pathway.

2. Materials and methods

2.1. Reagents

Dulbecco's Modified Eagle's Medium (DMEM), trypsin–EDTA and streptomycin ampicillin–amphotericin B mix, were purchased from HiMedia Pvt Ltd (Mumbai, India). Fetal bovine serum (FBS) was from Gibco (Grand Island, NY). 3-Isobutyl-1-methylxanthine (IBMX), dexamethasone, insulin, GW9662, wortmannin, protease inhibitor cocktail, rosiglitazone (RG), 2-deoxy-2-[7-nitro-2,1,3-benzoxadiazol-4-yl]amino]-D-glucose (NBDG) and 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT) were from Sigma-Aldrich (U.S.A). Adiponectin ELISA kit was from Cayman Chemical (U.S.A). Monoclonal anti-GLUT4 antibody, secondary anti-mouse immunoglobulin (IgG; Fab specific) – fluorescein isothiocyanate (FITC) antibody-mouse (sc-2078), mouse monoclonal antibodies for PPARγ (sc7273), GLUT4 (sc53566), IRS-1 (sc-559), phospho-IRS-1^{Tyr 632} (sc-17196), Akt (sc-5298), phospho-Akt^Ser473 (sc-33437), β-actin (sc-8432) and bovine anti-mouse IgG-HRP secondary antibody (sc-2371) were from Santa Cruz Biotechnology (U.S.A) Lanthascreen TR-FRET PPARγ competitive binding assay kit was from Invitrogen (U.S.A). cDNA synthesis kit and SYBR Green master mix was from Thermo Fisher Scientific (U.S.A).

2.2. Isolation of Licarin B

Myristica fragrans mace (500 g) was dried in an air oven maintained at 50 °C for 1 h, and then powdered mechanically. The powdered material was subjected to extraction with acetone (3 × 3 L) at room temperature. 120 g of acetone extract was obtained after removing the solvent. The acetone extract (120 g) was fractionated using column chromatography on silica gel (100–200 mesh) by eluting with hexane, followed by 5% ethyl acetate in hexane and finally with acetone. The acetone eluted portion (67.17 g) was subjected to column chromatography on silica gel (100–200 mesh) and eluted with hexane–ethyl acetate mixtures of increasing polarities to give 31 fraction pools based on the similarities of TLC. Fraction pool 2 (5.7 g) was subjected to column chromatography on silica gel using 5% ethyl acetate in hexane which yielded LB (600 mg). The structure of the compound was then confirmed by H-NMR and C-NMR as described elsewhere (Cao et al., 2013). Purity was determined to be ≥99%. The stock solution of LB was diluted in 0.05% DMSO (stock final concentration, 10 mM) for biological testing.

2.3. MTT assay

Cytotoxicity of LB in 3T3-L1 cells was measured by means of MTT assay. Briefly, cells after incubation with the compound (1–500 μM) for 48 h, were washed and MTT (0.5 g l⁻¹), dissolved in PBS, was added to each well for the estimation of mitochondrial dehydrogenase activity as described previously by Mosmann (1983).¹⁸ After 4 h of incubation at 37 °C in a CO₂ incubator, 10% SDS in DMSO was added to each well and the absorbance of solubilised MTT formazan products were measured at 570 nm after 45 min, using a microplate reader (Bioteck, U.S.A). Results were expressed as percentage of cytotoxicity. Based on viability data the concentrations of LB (5, 10 and 15 μM) have been selected for following experiments.

2.4. 3T3-L1 cell culture and adipocyte differentiation

For standard adipocyte differentiation, pre-adipocyte (3T3-L1; ATCC CL-173) cells were plated at a density of 1 × 10⁶ cells per well in 6-well plates and were grown to confluence in DMEM with 10% calf serum and 2 days post-confluence differentiation was induced by DMEM with 10% fetal bovine serum (FBS), 1 μM dexamethasone (DEX), 0.5 mM methylisobutylxanthine (IBMX) and 100 nM insulin with either vehicle (0.1% DMSO), positive control (10 μM rosiglitazone), or LB in DMSO (5 μM, 10 μM, 15 μM). From day 4 to the end of the experiment at day 8, the cells were re-fed every second day with DMEM with 10% fetal bovine serum (FBS), 100 nM insulin and with either vehicle (0.1% DMSO), positive control (10 μM rosiglitazone), or LB in DMSO (5, 10, 15 μM).¹⁹ GW9662 (10 μM), an antagonist of PPARγ, was added to cells along with the higher concentration of LB (15 μM) throughout the process of differentiation to confirm the PPARγ mediated effect of LB on adipocyte differentiation.

2.5. Oil-red-O staining

Differentiated adipocytes were stained with oil-red-O using the method of Kasturi and Joshi (1982) with some modifications.²⁰ Cells were washed twice with PBS and fixed with 70% ethanol for 30 min. Fixed cells were stained with oil-red-O solution (0.5% oil-red-O in isopropanol diluted with water (3 : 2)) followed by filtration for 1 h at room temperature and washed twice with distilled water. The images of stained cells were captured with a camera attached to a phase contrast microscope (Nikon Eclipse TS 100, Japan) at 40× magnification. For the quantification of triglyceride (TG) content in the cells, 100 μL of 100% isopropanol was added to each well. It was gently mixed for 10–15 min and the absorbance was read at 490 nm with a plate reader (Biotek, U.S.A).

2.6. LanthaScreen TR-FRET PPARγ competitive binding assay

A LanthaScreen TR-FRET PPARγ competitive binding assay was conducted according to the manufacturer's protocol (Invitrogen, U.S.A). Briefly, when a fluorescent ligand (tracer) is bound to the receptor (the conjugate of antibody and nuclear receptor protein), energy transfer from the antibody to the tracer occurs, and a high 520/495 ratio is detected. A compound under test displaces the tracer from PPARγ-LBD (ligand binding domain), which leads to a decrease in the FRET signal and the observation of a low TR-FRET ratio.²¹ The assay was performed in black non-coated, low volume and round bottomed 384-well corning plates. The compounds were diluted in TR-FRET assay buffer and 10 μl per well of these dilutions was dispensed in quadruplicates. A mixture of 5 nM of GST (glutathione S-transferase)-tagged PPARγ ligand binding domain (GST-PPARγ-LBD), 5 nM Tb ant-GST-antibody, 5 nM Fluormone Pan-PPAR Green, and serial dilutions of LB, GW9662 or RG were mixed and incubated in the dark for 2 h. The TR-FRET signals were measured by excitation at 340 nm and emission at 520 nm for fluorescein and excitation at 340 nm and emission at 490 nm for terbium in a Tecan multiplate reader (Tecan Infinite 200PRO, Mannedorf, Switzerland). For the TR-FRET ratios, the emission signal at 520 nm was divided by the emission signal at 495 nm and graph was plotted against the concentrations of LB to obtain the IC₅₀ value.

2.7. Molecular docking

Docking of LB into the PPARγ ligand binding domain was done using the softwares Autodock 4.2 and iGEMDOCK.^22,23 These docking softwares are used to find the appropriate binding and conformations of the ligand to the receptor. The 3D model of PPARγ was retrieved from the Brookhaven Protein Data Bank (http://www.rcsb.org/pdb; PDB ID 2Q5S). LB (ID 219296) and structure was downloaded from Chemspider (http://www.chemspider.com) and converted to pdb file using Chem3D Pro 10. The docking fitness of the ligand molecules to PPARγ and the amino acids of receptor (PPARγ) involved in interaction was predicted using the software iGEMDOCK. The binding energy of the ligand-PPARγ was analyzed using the software Autodock 4.2.

2.8. Western blot analysis

3T3-L1 cells were plated at a density of 1 × 10⁶ cells per well in 6-well plates and were subjected to differentiation. Then cells were treated with vehicle (DMSO), positive control (10 μM RG) or test compounds (5, 10 and 15 μM) for 8 days. At the end of treatment, cells were washed twice with ice cold PBS and lysed in ice-cold lysis buffer (50 mM Tris, 150 mM sodium chloride, 1.0% Triton X-100 and protease inhibitor cocktail, pH 8.0). The protein content of the lysate was measured with BCA protein assay kit (Thermo Scientific, U.S.A) using BSA as the standard. The lysate containing 30 μg of protein was subjected to SDS-PAGE on 10% gel and transferred onto nitrocellulose membrane by wet blotting. The membrane was incubated for 1 h at room temperature in 5% BSA in Tris-Buffered Saline (TBS) as blocking buffer, washed three times with TBS with Tween 20 (TBST-50 mM Tris–HCl, pH 7.4, 150 mM NaCl, 0.05% Tween-20) and probed overnight at 4 °C with primary antibody against PPARγ, GLUT4, IRS-1, phospho-IRS-1 (Tyr632), Akt, phospho-Akt (Ser473) or β-Actin (1 : 500 dilution). After washing three times in TBST for 10 min each, the membrane was incubated with horse radish peroxidase (HRP) conjugated secondary antibody (bovine anti-mouse IgG HRP, 1 : 500 dilution) for 2 h. Bound antibodies were detected using an enhanced chemiluminescence substrate (Bio-Rad, U.S.A) and measured by densitometry using a Chemi Doc XRS digital imaging system and the Multi Analyst software from Bio-Rad Laboratories (U.S.A).

2.9. Adiponectin estimation

The supernatant was collected on the 8^th day of differentiation from various experimental groups and the quantity of adiponectin secreted was estimated by a solid-phase sandwich ELISA kit (Cayman chemicals, U.S.A) and the concentrations were calculated according to the manufacturer's instructions.

2.10. Total RNA isolation and real time quantitative PCR

The mRNA level expression of CCAAT/enhancer-binding protein alpha (C/EBPα), insulin receptor substrate 2 (IRS-2) and lipoprotein lipase (LPL) was determined using quantitative real time PCR. Total RNA was extracted from differentiated 3T3-L1 cells using Trizol Reagent (Sigma-Aldrich, U.S.A). cDNA was generated using the cDNA synthesis kit (Thermo Scientific, U.S.A). The sequences of forward and reverse primers of C/EBPα, IRS-2 and LPL are listed in Table 1. All PCRs were carried out using SYBR Green master mix (Thermo Scientific, U.S.A) and amplification of each gene was performed on a C1000 Thermal cycler (Bio-Rad, U.S.A). The real-time cycling conditions were: initial enzyme activation at 50 °C for 2 min, denaturation at 95 °C for 10 min followed by 30 cycles of denaturation at 95 °C for 15 s and annealing/extension at 60 °C for 1 min. The product purity was confirmed by a dissociation curve analysis. The mRNA levels of the target genes were normalized to the values of β-actin, and presented as fold changes relative to controls. The fold change in the target gene relative to the β-actin endogenous control gene was determined by:

Fold change = 2^−Δ[ΔC_T]

where ΔC_T = C_T,target − C_T,β-actin and Δ[ΔC_T] = ΔC_T,stimulated − ΔC_T,control. C_T [threshold cycle] is the intersection between an amplification curve and a threshold line.

Table 1 Oligonucleotide primer sequences for qRT-PCR. C/EBPα – CCAAT/enhancer-binding protein alpha, IRS-2 – insulin receptor substrate 2 and LPL – lipoprotein lipase

Gene	Primer	Sequence
β-Actin	Sense	5′-AGTACCCCATTGAACGC-3′
	Antisense	5′-TGTCAGCAATGCCTGGGTAC-3′
C/EBPα	Sense	5′-AGACATCAGCGCCTACATCG-3′
	Antisense	5′-TGTAGGTGCATGGTGGTCTG-3′
LPL	Sense	5′-GAGATTTCTCTGTATGCCACC-3′
	Antisense	5′-CTGCAAATGAGACACTTTCTC-3′
IRS2	Sense	5′-GGCCTCTGTGGAAAATGTCTC-3′
	Antisense	5′-CTGTGGCTTCCTTCAAGTGAT-3′

---

2.11. Immunofluorescence assay

Fully differentiated 3T3-L1 adipocytes were treated with different concentrations of LB (5, 10, 15 μM) for 24 h. Cells were then insulin sensitized for 10 min (100 nM) and then fixed in 4% paraformaldehyde and 0.02% Triton X-100 in PBS. All subsequent steps in the whole-cell immunofluorescence labeling were done at room temperature. Fixed cells were rinsed with PBS three times and incubated with a polyclonal GLUT4 antibody for 1 h. Cells were washed three times with PBS for 5 min each and incubated with FITC conjugated secondary antibody conjugated to Texas red for 1 h. The cells were washed and imaged using spinning disc fluorescence microscope (BD Biosciences, U.S.A).

2.12. Glucose uptake by 2-NBDG assay

Fully differentiated 3T3-L1 adipocytes were treated with different concentrations of LB (5, 10, 15 μM) and RG (10 μM) for 24 h. Cells were then incubated in a low glucose medium in the presence of different concentrations of LB and RG for 3 h. This was removed after 3 h and replaced with media containing 100 nM insulin for 10 min to stimulate the cells. Following this, the cells were treated with 100 μM 2-NBDG²⁴ and incubated for 1 h. The cells were then washed twice with cold PBS and the fluorescence in cells was analyzed by FACS Aria™ II Flow cytometer (BD Bioscience, U.S.A).

2.13. Statistical analysis

All data represent the means ± SD of at least six individual experiments unless otherwise indicated in the text. Data were analyzed using SPSS v9.0 software (SPSS Inc., Chicago, IL, U.S.A). Replicates were averaged before entry as a single data point. Statistical significance was determined using one way ANOVA with significance accepted at p < 0.05. If F reaches significance, the Duncan's post hoc test was used to compare groups.

3. Results and discussion

3.1. Cytotoxicity of LB in 3T3-L1 preadipocytes

MTT assay was performed to assess the cytotoxicity of LB. The compound did not cause significant cell death up to 500 μM concentration (Fig. 1B). This reveals the potential of this compound for the development of a nutraceutical if it is scientifically validated for nutritional and medicinal use.

Fig. 1 (A) Structure of LB (B) effect of LB on 3T3-L1 cell viability. 3T3-L1 cells were treated with various doses (1–500 μM) of LB. Cell viability was measured by MTT assay after 48 h. Values are expressed as mean ± SD where n = 6. * indicates significant difference compared to vehicle control p < 0.05; MTT assay: methyl thiazol tetrazolium assay.

3.2. LB moderately increases TG accumulation during adipogenic differentiation and showed partial PPARγ agonist activity in 3T3-L1 preadipocytes

PPARγ is considered the master regulator of adipocyte differentiation and drives the expression of adipocyte-specific genes.^25,26 Together with C/EBPα, PPARγ activates terminal differentiation by transactivating the expression of downstream adipocyte-specific genes including adipocyte protein 2 (aP2), lipoprotein lipase (LPL), as well as adiponectin which facilitates the cytoplasmic storage of massive amounts of TGs.²⁷ In the present study, we analyzed the effect of LB on TG accumulation during adipogenesis. LB enhances adipocyte differentiation in 3T3-L1 cells through activating PPARγ. As shown in Fig. 2, LB at 15 μM concentration significantly promoted adipocyte differentiation and caused intracellular TG accumulation. However, compared to RG, the potential of LB was less on promoting adipocyte differentiation and increasing intracellular TG accumulation (Fig. 2A and B). These results suggest that LB causes differentiation of 3T3-L1 preadipocytes, but the intensity is weaker than RG. To confirm LB as a PPARγ ligand, we examined the effect of GW9662 in these experiments; co-treatment of LB with GW9662 resulted in a significant inhibition of 3T3-L1 preadipocyte differentiation (Fig. 2A) and TG accumulation (Fig. 2B). Overall these findings suggest that LB enhanced PPARγ activity and caused 3T3-L1 preadipocyte differentiation with the intensity significantly lower than RG. Also, co-treatment with GW9662, a PPARγ specific antagonist, abolished the effect of LB on 3T3-L1 preadipocyte differentiation, indicating that LB acts through PPARγ path way.

Fig. 2 LB promotes preadipocyte differentiation by activating PPARγ: two-day post confluent 3T3-L1 preadipocytes (day 0) were induced to differentiate with 5 μM, 10 μM, 15 μM LB or 10 μM RG and was replaced every 2 days along with the relevant media cocktail up to day 8. The assays were performed on day 8. (A) Representative microscopic images of differentiated 3T3-L1 adipocytes treated with various concentrations of LB, RG and LB + GW9662 (original magnification 40×). (a) Vehicle control (vc) (b) 10 μM RG (c) 5 μM LB (d) 10 μM LB (e) 15 μM LB. (f) 15 μM LB + 10 μM GW9662. (B) Representative microscopic images of differentiated 3T3-L1 cells stained with oil-red-O (a) vehicle control (b) 10 μM RG (c) 5 μM LB (d) 10 μM LB (e) 15 μM LB treated (f) 15 μM LB + 10 μM GW9662. Scale bars correspond to 100 μM. (C) Absorbance was spectrophotometrically determined at 490 nm after oil-red-O staining. Treatment groups are C-vehicle control, 10 μM RG, 5 μM LB, 10 μM LB, 15 μM LB treated, 15 μM LB + 10 μM GW9662. Results are mean ± SD (n = 6). (D) Representative western blot analysis of PPARγ and β-actin showing separate bands for PPARγ1 and PPARγ2 (E) the relative intensity of each band was quantified with β-actin. Data are presented as mean ± SD; n = 6; * indicates significant difference compared to vehicle control, ** indicates significant difference compared to rosiglitazone treated group-p < 0.05.

In adipocytes, two PPARγ isoforms exist, PPARγ1 and PPARγ2 differing only in their amino-terminal A/B domain. PPARγ2 contains an additional 30 amino acids and is specifically expressed in adipocytes, being the essential regulator of adipogenesis.²⁶ As the DNA-binding activity of PPARγ might be dependent on the abundance of the protein,²⁸ we performed western blot analysis of whole-cell lysate from differentiated cells incubated with or without LB or RG. PPARγ appears as a double band; the upper and the lower band corresponding to PPARγ2 and PPARγ1 isoforms, respectively. Expression of PPARγ1 and PPARγ2 were detectable in undifferentiated cells but found to increase considerably upon adipogenic differentiation. The results showed a decrease in the abundance of PPARγ2 isoform in the cells incubated with LB at different doses compared to RG treated group (Fig. 2D and E). These findings support the partial agonist property of LB in addition to the result from adipocyte differentiation experiments.

3.3. Competitive binding affinity of LB to PPAR γ

In vitro ligand/receptor interaction of LB to PPARγ receptor was analyzed by a Lanthascreen time-resolved fluorescent energy transfer (TR-FRET) PPARγ competitive binding assay²⁹ using RG and GW9662 as controls. The reaction of this assay includes the binding of the test compounds (RG, LB, and GW9662) with PPARγ LBD by displacing the Fluormone Pan PPAR Green which causes a reduction in the ratiometric emission at 520 nm/490 nm (520/490). This ratiometric emission at 520/490 nm was plotted against various concentrations of the ligands. The IC₅₀ value of the compounds for 50% displacement of Pan PPAR Green was calculated from the response curve and is found to be 2.4 μM for LB, whereas for RG and GW9662 it was 57.96 nM and 18.68 nM, respectively (Fig. 3). Thus the results of the competitive binding assay point out that LB, binds to the LBD of PPARγ predicting the agonist property of LB. However, the binding affinity was less compared to full agonist RG and unlike GW9662, LB does not act as a PPARγ antagonist. Thus, its properties are likely to be related to a partial agonist.^29,30

Fig. 3 LB binds to PPARγ LBD as evident from PPARγ competitive binding assay. To verify direct binding of LB to PPARγ, a Lanthascreen TR-FRET PPARγ competitive binding assay was performed. PPARγ agonist RG and PPARγ antagonist GW9662 (0–10 μM), were used as comparative controls. The reaction mixture contained 0.5 nm PPARγ-LBD (GST), 5 nm terbium-tagged anti-GST antibody, 5 nm Fluormone Pan-PPAR Green, 5 mm dithiothreitol, and varying concentrations of LB (0–10 μM) and RG (0–10 μM). After 3 h incubation in the dark, TR-FRET measurements were made in a Tecan multiplate reader. Data are mean ± SD where n = 6. Statistical significance is indicated by *, p < 0.05 compared with the TR-FRET ratio of vehicle.

3.4. Docking analysis of PPARγ with LB

Our molecular docking experiment using Autodock 4.2 and iGEMDOCK v2.1, confirmed a proper binding and conformation of LB to the LBD of PPARγ. In the molecular docking analysis, LB was predicted to bind at the ligand binding site of PPARγ (PDB entry 2Q5S). Autodock 4.2 has estimated the free binding energy of LB to be −9.36 kcal mol⁻¹ showing moderate binding affinity to the binding site (Table 2). The best docking pose of LB, as obtained from Autodock 4.2, is shown in Fig. 4A. The docking mode of LB with PPARγ, as obtained from iGEMDOCKv2, is shown in Fig. 4B. According to the binding model, LB also forms hydrophobic interactions with PHE-264, HIS-266, ILE-281, GLY-284, CYS-285, ARG-288, ILE-341 and SER-342 (Fig. 4A and B and Table 2). The interactions between PPARγ and LB are shown in the interaction table (Table 2) as obtained from iGEMDOCKv2.1. LB adopts a distinct binding mode and has no H-bonding interactions with PPARγ. Previous studies have shown that full agonists, such as RG, form conserved H bonds with the activation function helix (AF-2 helix) which, in turn, enhances the recruitment of co-activators.³¹ Thus in silico docking studies confirmed that LB adopts a distinct binding mode and have no H-bonding interactions with PPARγ. The absence of H-bonding interaction with the protein provides an explanation why LB functions as a partial agonist since most of the full agonists form a conserved H-bond with the AF-2 helix which, in turn, increases the recruitment of co-activators.³¹

Table 2 (A) Final genetic algorithm of the PPARγ-LB docked state: electrostatic energy is calculated using the software Autodock 4.0. (B) Interaction table and fitness scores of LB from iGEMDOCKv2.1. Table depicts the key amino acid residues of PPARγ that interact with LB and their fitness score in the post-screening analysis. In the table, energy represents the binding energy; vdW – van der Waals interaction H-bond – hydrogen bond

A
Estimated free energy of binding (ΔG^0)	−9.36 kcal mol^−1
Estimated inhibition constant (Ki)	137.56 nM at 298.15 K
Final intermolecular energy	−10.85 kcal mol^−1
vdW + H bond + desolv energy	−10.83 kcal mol^−1
Electrostatic energy	−0.02 kcal mol^−1
Final total internal energy	−0.87 kcal mol^−1
Torsional free energy	+1.49 kcal mol^−1

---

B
Amino acid residues interacting with Licarin B	Fitness score
PHE-264	−12.653
HIS-266	−6.3942
ILE-281	−8.3538
ILE-281	−6.6687
GLY-284	−7.08
CYS-285	−7.3264
ARG-288	−9.1028
ILE-341	−6.9065
ILE-341	−10.509
SER-342	−4.0839

---

Fig. 4 Docking simulation was performed to identify interaction between PPARγ and LB. (A) and (B) Best docking poses of LB on human PPARγ using (A) Autodock 4.2 and (B) iGEMDOCKv2.

3.5. LB modulates the pattern of expression of adipogenic genes during the differentiation of 3T3-L1 preadipocytes

To determine the effect of LB on the pattern of expression of PPARγ target genes which are critical for the expression of the adipocyte phenotype and insulin sensitivity, we analyzed mRNA expression of genes IRS-2, LPL, and C/EBPα in differentiated 3T3-L1 adipocytes. Exposure of a confluent population of 3T3-L1 preadipocytes to the adipogenic inducers (DEX, MIX, insulin, and FBS) along with different concentrations of LB (5, 10, 15 μM) or RG (10 μM) resulted in an increase in the expression of these adipogenic genes. Fig. 5 shows that treatment with insulin alone does not change the expression of these PPARγ target genes. The addition of LB (15 μM) increased C/EBPα, IRS-2 and LPL mRNA expression by 6.2, 7.6 and 7.02 fold respectively, relative to the expression of β-actin. Meanwhile, RG increased C/EBPα, IRS-2 and LPL mRNA expression by 14.02, 11.1, 12.9 fold respectively. Thus our results indicate that although LB failed to increase PPARγ gene expression like a full agonist RG, it moderately increases the expression of transcriptional target genes of PPARγ, such as IRS-2, LPL and C/EBPα upregulation of PPARγ target genes by LB could be related to the ability of the same to transactivate PPARγ as a ligand.

Fig. 5 LB modulates mRNA expression levels of PPARγ target genes c/EBPα, IRS-2 and LPL in differentiated 3T3-L1 adipocytes. 3T3-L1 adipocytes were differentiated along with indicated concentrations of LB (5, 10 and 15 μM) or RG (10 μM) up to day 8. After treatment, qRT-PCR was performed on isolated total RNA. The mRNA results are expressed as difference in fold change in treated cells compared to vehicle control. Data are mean ± SD values where n = 6. *, **, p < 0.05 versus control.

3.6. LB induced adiponectin secretion in differentiated 3T3-L1 adipocytes

Adiponectin is one of the major adipokines exclusively secreted from adipose tissue, which plays a significant role in increased glucose uptake and fatty acid oxidation, and has been shown to improve insulin sensitivity.³² In the present study fully differentiated 3T3-L1 cells were treated with different concentrations of LB or RG for 24 h, and the release of adiponectin from the cells into the medium was determined using a mouse adiponectin ELISA kit. The quantitative ELISA analysis demonstrated that adiponectin concentrations in the conditioned medium significantly increased (p < 0.05) 24 h after LB or RG treatment (Fig. 6) in 3T3-L1 adipocytes. Results showed that LB increased adiponectin secretion in a dose-dependent manner. Consistent with previous reports, we also observed that the PPARγ agonist RG enhanced adiponectin secretion by 47.8%, while LB at concentrations of 5, 10 and 15 μM increased the same by 20.9, 29.96 and 39.52% respectively compared to control.³³ This data suggests that increased secretion of adiponectin by the treatment of LB plays a role in alleviating insulin resistance. This observation again supports partial agonistic feature of LB. Altogether it suggests that LB influenced the expression of LPL, adiponectin secretion, decreased fatty acid utilization and triglyceride synthesis in 3T3-L1 cells. This, in turn, affected lipid accumulation during adipocyte differentiation through the partial suppression of PPARγ activity.

Fig. 6 LB upregulates adiponectin secretion in 3T3-L1 cells. Cells were differentiated and treated as described earlier up to day 8. Adiponectin levels in the conditioned medium on the 8^th day were measured by ELISA. Data are presented as mean ± SD * indicates significant difference compared to vehicle control, ** indicates significant difference compared to RG treated group (p < 0.05) where n = 6.

3.7. LB enhances the expression and translocation of the insulin-responsive glucose transporter GLUT4 in differentiated 3T3-L1 preadipocytes

Insulin-stimulated GLUT4 translocation is an important event in glucose uptake in adipocytes. So in the study, we analyzed the effect LB on GLUT4 expression and translocation in 3T3-L1 adipocytes. Fully differentiated 3T3-L1 adipocytes were probed with anti-GLUT4 antibodies followed by fluorescent detection (Fig. 7A). Treatment of 3T3-L1 cells with insulin produced an increase in GLUT4 at the plasma membrane consistent with previously published data.³⁴ We noticed that the GLUT4 fluorescence intensity increased in the plasma membrane of LB treated cells, than the control group, showing that LB significantly induced GLUT4 translocation in 3T3-L1 adipocytes (Fig. 7A and B). Interestingly, the enhancement of insulin-induced GLUT4 translocation by LB was stronger than that of RG (Fig. 7B).

Fig. 7 LB enhances GLUT4 translocation and expression in 3T3-L1 cells. GLUT4 expression is determined by western blot and immunofluorescence with antibody against GLUT4. (A) Translocation of GLUT4 protein was measured through immunofluorescence and was analyzed by fluorescent microscope imaging (original magnification 20×). (a) Vehicle control (b) 10 μM RG (c) 5 μM LB (d) 10 μM LB (e) 15 μM LB. Scale bars correspond to 100 μM. (B) Bar diagram showing the relative fluorescence intensity representing changes in GLUT4 expression in different groups (C) representative western blot analysis of GLUT4 and β-actin (D) the relative intensity of each band of GLUT4 protein to its own β-actin was quantified. Data are presented as mean ± SD; (n = 6) * indicates significant difference compared to control, ** indicates significant difference compared to RG treated group – p < 0.05.

To investigate whether insulin alone or insulin with LB increased GLUT4 protein expression, western blot was performed. In western blots of whole-cell lysates, expression of the total cellular GLUT4 was detectable in undifferentiated cells but considerably increased upon differentiation (Fig. 7C). Expression of GLUT4 was pronounced in LB treated cells in a dose-dependent manner compared to RG treated cells (Fig. 7C and D). Taken together these results suggest that LB induced GLUT4 expression as well as translocation similar to RG in 3T3-L1 adipocytes.

3.8. LB enhances insulin signaling via IRS-1/PI3K/AKT pathway in differentiated 3T3-L1 cells

The insulin signaling via IRS-1/PI3K/AKT pathway has been established as an upstream of GLUT4 protein. Under physiological conditions phosphorylation of PI3K is promoted by IRS-1, which is a proximal substrate of the insulin receptor. PI3K subsequently phosphorylates AKT and promotes GLUT4 translocation to the membrane from inner vesicles and consequently stimulates glucose uptake.³⁵ Previous studies have shown that a decreased insulin-stimulated glucose uptake is associated with reduced tyrosine phosphorylation of IRS-1 and PI3K activation in an insulin-resistant state.³⁶ Therefore to investigate the molecular mechanisms underlying the improved glucose transport by LB, we evaluated the effect of LB on phosphorylation of IRS-1 and Akt, two important factors in the insulin signaling pathway in differentiated 3T3-L1 adipocytes. Consistent with the result of glucose transport and GLUT4 translocation (Fig. 8), treatment of cells with 5, 10 or 15 μM LB improved Tyr632 phosphorylation of IRS-1 and Ser473 phosphorylation of Akt (Fig. 8). As shown in Fig. 8C and D, 10 μM wortmannin, a specific inhibitor of PI3K, suppressed the increase of Akt phosphorylation induced by 15 μM LB in differentiated 3T3-L1 cells. This result indicates the effect of LB on of Akt phosphorylation is PI3K dependent. Taken together results of the present study suggest that LB improved GLUT4 translocation and glucose transport in differentiated 3T3-L1 cells by the activating the insulin signaling pathway via phosphorylation of IRS1 (Tyr 632), PI3K activation, and phosphorylation of Akt (Ser473). Interestingly the magnitude of the effect of LB on insulin signaling pathway was comparable to that of full agonist RG.

Fig. 8 LB enhances insulin signaling via IRS1/PI3K/AKT pathway in differentiated 3T3-L1 cells. 3T3-L1 adipocytes were differentiated in the presence of RG or LB for 8 days. Cell lysates were subjected to western blot analysis. (A) Representative western blot analysis of IRS-1, p-IRS-1, AKT, p-AKT and β-actin (B) the band intensities of phospho-IRS-1 (pIRS-1) or pAKT were normalized by their total IRS-1 or AKT, respectively (C) the increase of AKT phosphorylation by LB in differentiated 3T3-L1 was suppressed by 10 μM wortmannin, a specific inhibitor of phosphatidylinositol-3-kinase (PI3K). 3T3-L1 cells were allowed to differentiate with LB (15 μM) in the presence or absence of wortmannin (10 μM) and cells were harvested for western blotting. Representative western blot analysis of AKT, p-AKT and β-actin. (D) The band intensities of pAKT were normalized by their total AKT. Data are presented as mean ± SD; (n = 6) * indicates significant difference compared to control, ** indicates significant difference compared to RG treated group. p < 0.05.

3.9. LB increased glucose uptake in differentiated 3T3-L1 adipocytes

The characteristic of insulin resistance is the decrease in insulin sensitivity and glucose uptake in adipocytes. Therefore, we examined the effects of LB on glucose uptake in differentiated 3T3-L1 adipocytes. The insulin-stimulated glucose uptake in 3T3-L1 cells was measured by flow cytometry using 2-NBDG, a fluorescent analogue of glucose. As shown in Fig. 9 LB could improve insulin-stimulated glucose uptake in differentiated 3T3-L1 adipocytes when compared to vehicle control. The results showed that treatment with LB at 5, 10, 15 μM concentrations increased glucose uptake by 9.1, 14.5, 26.5 fold respectively while the positive control RG caused a fold increase of 17.7 (Table 3).

Fig. 9 Measurement of glucose uptake using 2-NBDG in 3T3-L1 cells by flow cytometry. The representative histogram shows mean fluorescence level (530 nm emission), which indicates the glucose uptake activity. FITC histograms in (A) control, (B) RG (C) LB 5 μM, (D) LB 10 μM and (E) LB 15 μM.

Table 3 Effect of LB on insulin-stimulated glucose uptake in 3T3-L1 cells

Compound	Mean fluorescence level^a (%)	Fold increase^b
a Values are mean ± SD where n = 6; * indicates significant difference compared to control, ** indicates significant difference compared to RG treated group (p < 0.05).b Fold increase was calculated in comparison to control.c Positive control.
Control	5.8 ± 0.35
RG (10 μM)^c	23.5 ± 0.56*	17.7
LB (5 μM)	14.9 ± 0.26**	9.1
LB (10 μM)	20.3 ± 0.35**	14.5
LB (15 μM)	32.3 ± 0.98**	26.5

---

The synthetic ligands of PPARγ like the thiazolidinediones increases glucose uptake by enhancing the expression of GLUT1 or GLUT4 or by increasing the translocation of GLUT4.³⁷ It is also reported that PPARγ agonists can enhance the insulin sensitivity in adipocytes by increasing the number of insulin sensitive adipocytes or by increasing the expression of genes like adiponectin and GLUT4.^38,39 In a mature 3T3-L1 cell system, LB significantly improved insulin-stimulated glucose uptake, similar to a PPARγ full agonist, such as RG. This provides data to support the strong insulin sensitizing property of LB, similar to RG.

4. Conclusion

The biochemical, cell-based, and in silico data confirmed that LB acts as a partial PPARγ agonist, which preferentially activates adipocyte-specific genes including C/EBPα, IRS 2, lipoprotein lipase (LPL), as well as adiponectin and uncouples insulin signaling from TG accumulation in vitro. We showed that LB exhibits a partial agonist activity by modulating the expression of PPARγ target genes in 3T3-L1 cells and PPARγ dependent adipogenesis. LB upregulated GLUT4 translocation and expression via IRS-1/PI3K/AKT signaling pathway to a similar degree as that of RG implying gene-specific partial agonism. Taken together, these data indicate that LB is a compound with multiple targets that holds promising therapeutic potential in the treatment of T2DM with no toxicity. Moreover, Myristica fragrans mace is also used as an edible spice, has medicinal properties and a broad range of tolerance in the human body. This makes LB a potent new chemical entity in the development of nutraceuticals for diabetes especially for insulin resistance, after a detailed in vivo study.

Author contribution

Raghu KG conceived and designed the experiments. Shyni G. L. and Kavitha Sasidharan was responsible for the acquisition and interpretation of cell culture data and drafting of the manuscript. Sajin K. Francis and Mangalam S. Nair isolated and purified LB from Myristica fragrans. Arya A. Das conducted molecular docking studies.

Conflict of interest

The authors have declared no conflict of interest.

Abbreviations

TZDs	Thiazolidinediones
PPARγ	Peroxisome proliferator-activated receptor gamma
C/EBP-α	CCAAT/enhancer-binding protein alpha
T2DM	Type 2 diabetes mellitus
LPL	Lipoprotein lipase
IRS-2	Insulin receptor substrate 2
GLUT4	Glucose transporter 4
2-NBDG	2-Deoxy-2-[[7-nitro-2,1,3-benzoxadiazol-4-yl]amino]-D-glucose
IRS-1	Insulin receptor substrate 1

Acknowledgements

We are also thankful to the CSIR 12^th 5 year plan project ‘NaPAHA’ (CSC 0130) for partial financial assistance. We thank the Director, CSIR-NIIST and Head, Agroprocessing and Natural Products Division, CSIR-NIIST for providing necessary facilities. Dr Shyni G. L. is grateful to Department of Biotechnology (DBT, New Delhi) for the Postdoctoral fellowship.

References

1. M. de Groot, R. Anderson, K. E. Freedland, R. E. Clouse and P. J. Lustman, Psychosom. Med., 2001, 63, 619–630.
2. A. M. Jacobson, Endocr. Pract., 2004, 10, 502–508.
3. M. Simmonds and M. Howes, Plants used in the treatment of diabetes, in Traditional Medicines for Modern Time—Antidiabetic Plants, ed. A. Soumyanath, CRC Press/Taylor and Francis Group, 2006, vol. 6, pp. 19–82.
4. M. Bnouham, A. Ziyyat, H. Mekhfi, A. Tahri and A. Legssyer, Int. J. Diabetes Metab., 2010, 14, 1–25.
5. J. K. Grover, S. Yadav and V. Vats, J. Ethnopharmacol., 2002, 2002, 81–100.
6. J. Luo, D. M. Fort, T. J. Carlson, B. K. Noamesi, D. nii-Amon-Kotei, R. King, J. Tsai, J. Quan, C. Hobensack, P. Lapresca, N. Waldeck, C. D. Mendez, S. D. Jolad, D. E. Bierer and G. M. Reaven, Diabetic Med., 1998, 15, 367–374.
7. K. B. Christensen, A. Minet, H. Svenstrup, K. Grevsen, H. Zhang, E. Schrader, G. Rimbach, S. Wein, S. Wolffram, K. Kristiansen and L. P. Christensen, Phytother. Res., 2009, 23, 1316–1325.
8. J. Berger and D. E. Moller, Annu. Rev. Med., 2002, 53, 409–435.
9. B. G. Shearer and A. N. Billin, Biochim. Biophys. Acta, 2007, 1771, 1082–1089.
10. M. Ahmadian, J. M. Suh, C. Liddle, A. R. Atkins, M. Downes and R. M. Evans, Nat. Med., 2013, 19, 557–566.
11. S. Rocchi, F. Picard, J. Vamecq, L. Gelman, N. Potier, D. Zeyer, L. Dubuquoy, P. Bac, M. F. Champy, K. D. Plunket, L. M. Leesnitzer, S. G. Blanchard, P. Desreumaux, D. Moras, J. P. Renaud and J. Auwerx, Mol. Cell, 2001, 8, 737–747.
12. G. Y. Cao, X. W. Yang, W. Xu and F. Li, Food Chem. Toxicol., 2013, 62, 167–171.
13. A. Iyer, S. Panchal, H. Poudyal and F. Li, Indian J. Biochem. Biophys., 2009, 46, 467–481.
14. S. B. Patil, V. A. Ghadyale, V. A. Taklikar, C. R. Kulkarni and A. U. Arvindekar, Plant Foods Hum. Nutr., 2011, 66, 85–90.
15. A. E. Thomas and C. Wenying, Tetrahedron Lett., 1996, 37, 6969–6970.
16. E. D. Coy, L. E. Cuca and M. Sefkow, Bioorg. Med. Chem. Lett., 2009, 24, 6922–6925.
17. K. L. Han, J. S. Choi, J. Y. Lee, J. Song, M. K. Joe, M. H. Jung and J. K. Hwang, Diabetes, 2008, 57, 737–745.
18. T. Mosmann, J. Immunol. Methods, 1983, 65, 55–63.
19. J. B. Hansen, R. K. Petersen, B. M. Larsen, J. Bartkova and J. Alsner, J. Biol. Chem., 1999, 274, 2386–2393.
20. R. Kasturi and V. C. Joshi, J. Biol. Chem., 1982, 257, 12224–12230.
21. Y. Zhang, L. Yu, W. Cai, S. Fan, L. Feng, G. Ji and C. Huang, Sci. Rep., 2014, 4, 7375–7379.
22. K. C. Hsu, Y. F. Chen, S. R. Lin and J. M. Yang, BMC Bioinf., 2011, 12, 1–S33.
23. N. Mahindroo, C. C. Wang, C. C. Liao, C. F. Huang, I. L. Lu, T. W. Lien, Y. H. Peng, W. J. Huang, Y. T. Lin, M. C. Hsu, C. H. Lin, C. H. Tsai, J. T. Hsu, X. Chen, P. C. Lyu, Y. S. Chao, S. Y. Wu and H. P. Hsieh, J. Med. Chem., 2009, 49, 1212–1216.
24. M. C. Louzao, B. Espiña, M. R. Vieytes, F. V. Vega and J. A. Rubiolo, Glycoconjugate J., 2008, 25, 503–510.
25. Y. Barak, M. C. Nelson, E. S. Ong, Y. Z. Jones, P. Ruiz-Lozano, K. R. Chien, A. Kodar and R. M. Evans, PPAR gamma is required for placental, cardiac, and adipose tissue development, Mol. Cell, 1999, 4, 585–595.
26. E. D. Rosen and B. M. Spiegelman, J. Biol. Chem., 2001, 276, 37731–37734.
27. F. M. Gregoire, C. M. Smas and H. S. Sul, Physiol. Rev., 1998, 78, 783–809.
28. K. A. Burns and J. P. Vanden Heuvel, Biochim. Biophys. Acta, 2007, 1771, 952–960.
29. T. S. Hughes, P. K. Giri, I. M. de Vera, D. P. Marciano and D. S. Kuruvilla, et al., Nat. Commun., 2014, 5, 3571–3584.
30. G. L. Shyni, S. Kavitha, S. Indu, A. D. Arya, S. S. Anusree, V. P. Vineetha, S. Vandana, A. Sundaresan and K. G. Raghu, Biofactors, 2014, 646–657.
31. T. Waku, T. Shiraki, T. Oyama, Y. Fujimoto, K. Maebara, N. Kamiya, H. Jingami and K. Morikawa, J. Mol. Biol., 2009, 385, 188–199.
32. F. Yuchang, L. Nanlan, L. K. Richard and G. W. Timothy, J. Lipid Res., 2005, 46, 1369–1379.
33. S. S. Anusree, A. Priyanka, V. M. Nisha, A. A. Das and K. G. Raghu, Food Funct., 2014, 5, 2590–2601.
34. B. M. Spiegelman, Diabetes, 1998, 47, 507–514.
35. N. J. Bryant, R. Govers and D. E. James, Nat. Rev. Mol. Cell Biol., 2002, 3, 267–277.
36. P. Gual, Y. Le Marchand-Brustel and J. F. Tanti, Biochimie, 2005, 87, 99–109.
37. K. Claudia, E. A. Niki, H. Maria, H. Jana, H. Nicole, W. Jannette, L. S. Alexios and V. Gunter, Planta Med., 2011, 77, 346–353.
38. R. P. Brun and B. M. Spiegelman, J. Endocrinol., 1997, 155, 217–218.
39. L. Michael and A. L. Mitchell, Cell, 2005, 123, 993–999.

---

Footnote

† These authors contributed equally to this work.

---